Sub-iris Technique for Non-ideal Iris Recognition

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Iris recognition has become one of the most prominent biometrics due to its high recognition accuracy, and it is suitable for the verification and identification systems. Most iris recognition systems can recognize human identity especially when the iris region is free from noise. However, the iris region is typically occluded by noise, eyelid and eyelash. The occluded iris can hide rich iris features, which consists fewer iris features in the iris region, thus reducing the recognition accuracy of the system. The occluded iris is also called a non-ideal iris. In this study, an improvement is made to enhance the performance of an existing sub-iris technique. This can overcome the problem of obtaining iris features from the non-ideal iris. Moreover, the proposed method can also reduce the execution time of the iris recognition system. According to the results, the proposed method managed to localize the sub-iris region, while achieving an equal error rate of 3.75%; area under curve of 0.9832; execution time of 0.719 s; and decidability of 2.9332. These metrics are superior to those of other methods.

Original languageEnglish
Pages (from-to)7219-7228
Number of pages10
JournalArabian Journal for Science and Engineering
Volume43
Issue number12
DOIs
Publication statusPublished - 1 Dec 2018

Fingerprint

Computability and decidability
Biometrics
Identification (control systems)

Keywords

  • Biometrics
  • Execution time
  • Iris recognition
  • Non-ideal iris
  • Recognition accuracy
  • Sub-iris technique

ASJC Scopus subject areas

  • General

Cite this

Sub-iris Technique for Non-ideal Iris Recognition. / Jamaludin, Shahrizan; Zainal, Nasharuddin; Wan Zaki, Wan Mimi Diyana.

In: Arabian Journal for Science and Engineering, Vol. 43, No. 12, 01.12.2018, p. 7219-7228.

Research output: Contribution to journalArticle

@article{639e7e567d064d619b23fbe75c94fb7a,
title = "Sub-iris Technique for Non-ideal Iris Recognition",
abstract = "Iris recognition has become one of the most prominent biometrics due to its high recognition accuracy, and it is suitable for the verification and identification systems. Most iris recognition systems can recognize human identity especially when the iris region is free from noise. However, the iris region is typically occluded by noise, eyelid and eyelash. The occluded iris can hide rich iris features, which consists fewer iris features in the iris region, thus reducing the recognition accuracy of the system. The occluded iris is also called a non-ideal iris. In this study, an improvement is made to enhance the performance of an existing sub-iris technique. This can overcome the problem of obtaining iris features from the non-ideal iris. Moreover, the proposed method can also reduce the execution time of the iris recognition system. According to the results, the proposed method managed to localize the sub-iris region, while achieving an equal error rate of 3.75{\%}; area under curve of 0.9832; execution time of 0.719 s; and decidability of 2.9332. These metrics are superior to those of other methods.",
keywords = "Biometrics, Execution time, Iris recognition, Non-ideal iris, Recognition accuracy, Sub-iris technique",
author = "Shahrizan Jamaludin and Nasharuddin Zainal and {Wan Zaki}, {Wan Mimi Diyana}",
year = "2018",
month = "12",
day = "1",
doi = "10.1007/s13369-017-3051-8",
language = "English",
volume = "43",
pages = "7219--7228",
journal = "Arabian Journal for Science and Engineering",
issn = "1319-8025",
publisher = "King Fahd University of Petroleum and Minerals",
number = "12",

}

TY - JOUR

T1 - Sub-iris Technique for Non-ideal Iris Recognition

AU - Jamaludin, Shahrizan

AU - Zainal, Nasharuddin

AU - Wan Zaki, Wan Mimi Diyana

PY - 2018/12/1

Y1 - 2018/12/1

N2 - Iris recognition has become one of the most prominent biometrics due to its high recognition accuracy, and it is suitable for the verification and identification systems. Most iris recognition systems can recognize human identity especially when the iris region is free from noise. However, the iris region is typically occluded by noise, eyelid and eyelash. The occluded iris can hide rich iris features, which consists fewer iris features in the iris region, thus reducing the recognition accuracy of the system. The occluded iris is also called a non-ideal iris. In this study, an improvement is made to enhance the performance of an existing sub-iris technique. This can overcome the problem of obtaining iris features from the non-ideal iris. Moreover, the proposed method can also reduce the execution time of the iris recognition system. According to the results, the proposed method managed to localize the sub-iris region, while achieving an equal error rate of 3.75%; area under curve of 0.9832; execution time of 0.719 s; and decidability of 2.9332. These metrics are superior to those of other methods.

AB - Iris recognition has become one of the most prominent biometrics due to its high recognition accuracy, and it is suitable for the verification and identification systems. Most iris recognition systems can recognize human identity especially when the iris region is free from noise. However, the iris region is typically occluded by noise, eyelid and eyelash. The occluded iris can hide rich iris features, which consists fewer iris features in the iris region, thus reducing the recognition accuracy of the system. The occluded iris is also called a non-ideal iris. In this study, an improvement is made to enhance the performance of an existing sub-iris technique. This can overcome the problem of obtaining iris features from the non-ideal iris. Moreover, the proposed method can also reduce the execution time of the iris recognition system. According to the results, the proposed method managed to localize the sub-iris region, while achieving an equal error rate of 3.75%; area under curve of 0.9832; execution time of 0.719 s; and decidability of 2.9332. These metrics are superior to those of other methods.

KW - Biometrics

KW - Execution time

KW - Iris recognition

KW - Non-ideal iris

KW - Recognition accuracy

KW - Sub-iris technique

UR - http://www.scopus.com/inward/record.url?scp=85056268318&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85056268318&partnerID=8YFLogxK

U2 - 10.1007/s13369-017-3051-8

DO - 10.1007/s13369-017-3051-8

M3 - Article

VL - 43

SP - 7219

EP - 7228

JO - Arabian Journal for Science and Engineering

JF - Arabian Journal for Science and Engineering

SN - 1319-8025

IS - 12

ER -